Global expression profiles are excellent phenotypes. In cancer biology they can be used for diagnosis and prognosis. In yeast and other model organisms they have been used for discovering gene function and drug targets. In the previous funding period we found that expression profiles of Dictyostelium are also useful diagnostics. We also tested and proved the hypothesis that expression profiles can be used as phenotypes for genetic epistasis analysis and pathway construction. In this proposal we will assign function to hundreds of genes by using expression profiles as phenotypes to characterize mutants. We will also test the hypothesis that distinct patterns of global gene expression represent specific, definable biological functions. In a nutshell, we propose that changes in subsets of the expression pattern represent changes in specific biological functions such as cell-cell adhesion or chemotaxis. If this hypothesis were correct, one would be able to identify the specific biological functions that have been affected by mutations or drug treatment, turning the expression profile from an abstract phenotype into a direct diagnostic tool. The applicaiton of this discovery to human health is self evident as it would transform microarray diagnosis of cancer and other diseases from a comparative method into a direct diagnostic method. To reach these goals we will perform microarray analysis of gene expression on several hundred mutants, correlate their expression profiles with other phenotypes, such as chemotaxis, infer which biological functions are represented by which subset of the expression profile and test the functions directly by mutagenesis. In addition, we will analyze selected mutants for specific functions. We will test mutants in 20 of the 66 ABC transporters, all 19 bZIP transcription factors, mutants in a new chemotaxis pathway and mutants with altered sensitivity to the anti-cancer drug cisplatin. Successful completion of these goals will assign function to hundreds of new genes and turn expression profiling into a powerful tool for direct diagnosis of biological function. The data will also be disseminated to the community as a means of expanding the correlation between expression profiles and specific biological functions beyond the scope of this proposal.